Reprinted from

Journal ofPRAGMATICS

Journal of Piagmatits 27 (1997) 61-82

Interaction and causation:Causative constructions in modern Standard Dutch

Ane Verhageir1 *, Su/anne Kemmerb

Cenne foi Lani>uai>i and Commumcation Utieiht Umvci uty, Tians 10, 3512 JK UtiechtΊ hc Nethet laniJ<,

'' Dt/Mi tment of Lnii;untic\ RueUnnemty 6100 Main Stiect, Houston, TX 77005 1892 USA

Rcccivcd January 1996, icvised vcision February 1996

ELSEVIER

JOURNAL OF PRAGMATICS Λη/««;<//«;/>//««;\ nwnlhl\ o{ Ιαηι-ua^e \nidu\

Chict Idnoi Jacob L Mey, Odensc University Campusvej 55 DK-'OTO Odense M Dcnmark (e-mail jam@dou dk)Lilitui·! Claudia Cafli . Dipartimento di Scien?e Glottoetnologiche, Universita dl Genova, V i a Balbi 4, 1-16126 Genoa, Italy

(e-mail cambiaghi@pavia m f n u)Hartmut Haberland. Department öl Languages and Cu l tu i c . Universi ty ot Roskilde, PO Box 260, DK-4000Roskilde Denmark (e-mail hartmut@ruc dk)Richard W Janney, Ins t i tu t f ü r Englische Philologie, Universität München, Schellingstialie l, D-80799, Germany(e-mail janney@em um-frankfuit de)

ÄfUfM rdilon Rachel Giora Department of Poetics and Comparative Literature, Tel-Ανιν Univers i ty, Ramat A v i v IL-69978Tel-Aviv, Israel (e-mail giorar@ccsg tau ac i l )Marlene Dohtsky, Ins t i tu t Umversitaire de Formation des Maitrcs (IUFM), 10 lue Mohtor, F-V)OI6 Paris,France

Sc/uib Ldiloi Bruce Fräser, School öl Education, Boston Universi ty, 60S Commonwealth Avenuc Boston MA02211) USA (cmail bfraser@acs bu edu)

A\MHiate Iditon Linda Colcman University of MarylandFerenc Kieler Hunganan Academy öl SciencesJet Verschueren Belgian National Science FoundationJan-Ola Ostman, Universi ty of Hels inki

liuaiil <>l Aihiwn filiton l Allwood (Gothcnburg), H Arndl (Pulheim), S Blum Ku lka (Jerusalem), A Boguslawski(Warsaw) B Comnc (Los Angeles) F Coulmas ( lokyo) W U Dressler (Vienna) C Fillmore (Beikeley), T Frctheim (Trondheim), F Gregersen (Copenhagen) Υ Gu (Beijmg) E Hajicova (Piague) A Hakuhnen ( H e l s i n k i ) P Harder (Copenhagen)R M Harmsh (Tucson), R Hasan (Sidncy), H l laverkalc (Amsterdam), D Hymes (Charlottesvi l le), S Ide (Tokyo), A Kasher(Tel A v i v ) G Kasper ( H o n o l u l u ) R I akolf (Berkeley) A Luke (Brisbane) Yoshiko Matsumoto (Stanford), A McHoul(Murdoch). J Nuyts (Antwerp), H Parrel (Brüssels), T Pavhdou (rhessaloniki) U Quastholf (Dortmund) M Sbisa ( (ncs le)E Schegloff (Los Angeles), D S c h i f l n n (Washington, DC), D Tannen (Washington, DC) J Thomas (Lancaster), S Fhompson(Santa Barbara). J van der Auwera (Antwerp) , J Wagner (Odense), D Wilson (London)

Linguistic pragmatics has bcen able to formulate a nurnbcr of questions ovcr the ycars (hat are esscntial to oui understandingöl language äs people s mam Instrument öl natural and societal' mteraclion By providmg possiblc theoretical foundations loithc study of l inguis t ic practice l inguis t ic pragmatics has helped to extend our knowledge of the loims, iuncl ions andtoundations ot human intcraction The Inuinal of Ριαν,ηιιιΐιι s i d e n l i f i e s wi th the above general scope and aims of piagmatics

The Journal wclcomes authontat ive, innovat ive pragmalic scholarship from all practice oncnted h n g u i s t i c standpomts Itprovides a forum for pragmatic studies in sociolmguistics, gcncial l i n g u i s t i c s , conversation analysis discourse analysis, cogmtivehngui s t ic s computational l i n g u i s t i c s apphed l i n g u i s t i c s and other areas öl l inguist ic rcsearch In addition it endcavois tonarrow the distance belween l inguist ics and such neighbonng disciphnes äs cominumcation science (mcludmg the study of face-to-face interaction and nonverbal commumcation), Information science (mcludmg a r l i f i c i a l intelhgcnce rcsearch and in generalthe theory and practice of machme-human interaction) psychiatiy (mcludmg thc study ot schi/ophiemc spcech) andneuropsychology (mcludmg the study of speakmg and readmg disorders) The luuinal wclcomes both those contributionsonginatmg m linguistics proper and those takmg neighbonng related fields äs their pomt of departure

All manuscnpts and correspondence conceinmg cditonal matters, should be sent to the Chiel Editor Books lor ic-view andunsolicited book reviews should be sent to the Review Editor, books sent for revicw will not be icturncd All contributions,mcludmg reviews, are expected to comply with the nnln« ΐιυιΐΊ lo mulion pnnted on the mside hack cover of cach issue

'iiihunption mjoiinution louinul oj Ριαφηιιιιι\ (ISSN 0178-2166) For 1997 volumes 27-28 are scheduled tor publ ica t ionSubscnption prices are avai lable upon request from the puhhsher Subscriptions are acccpted on a prepaid basis only and aieentered on a calendar year basis Issucs are sent by sui facc mail except to the tol lowing countnes wherc air dchvery via SAL isensured Aigcntma, Austraha, Bra/il Canada Hong Kong, India Israel, Japan Malaysia, Mexico, New /ealand, Pakistan, PRChina Smgapore, South Atnca, South Koiea Taiwan Thailand USA For all other countnes a i rmai l rates are avai lable uponrequest Claims lor missmg issues must be made w i t h i n six months of our publication (maihng) date Please addrcss all youirequests regardmg Orders and subscnption quencs to Elscvier Science B V Customer Suppoi t Department PO Box 2l l , 1000AE Amsterdam fhe Nctherlands Fax +11 20 4ΗΉ412'suhunptwin ul i/w icil ι i'dm i'd ι tue hn imtnhti\ i>/ l/u ΙΙΊΑ Prcviously. the IPrA Secretanat has collectecl Orders and paymcntfor these reduced rate subscnptions f i o m i l s membcrs and subscquently sent thesc to Eisevier Science B V foi l u i t h e r handlmgIPrA members may order these subscnptions directly Irom Eisevier Science B V

Copyright © 1997 Elsevicr Science B V All righls reservedThis Journal and the m d i v i d u a l contr ibut ions conlamed in i t are protected by Ihe Copyright öl Eisevier Science B V and thefol lowmg terms and conditions apply lo their usePhotocopying. Smgle photocopies of smgle articles may be madc for personal usc äs allowed by nat ional Copyright lawsPermission of the pubhsher and payment öl a tce is rcquned toi all other photocopymg including mu l t i p l e or systematic copy-ing copying for advertising or promotional puiposes. resale and all torins öl document dchvery Spccial uites aie available loieducational msti tut ions that wish to make photocopies lor non prohl educational classroom use

In the U S A users may clear permissions and make payment through the Copynght Cleaiance Center, 222 Rosewood DuveDanver, MA 01923, USA In the UK users may clear permissions and make payment thiough thc Copyright Licensmg AgencyRapid Clearance Service (CLARCS) 90 Totlenham Court Road, London W1P OLP UK In other countnes whcic a local copynght clearance centre exisf , please contact it for In fo rmat ion on required permissions and paymentsDerivative Works. Subscribers may reproduce tables of contents or prepare hsts öl articles mcludmg abstacts lor mternal cir-culation wi th in their msti tut ions Permission of the pubhsher is rcquned loi icsale 01 dislnbution outside the msitution

Permission of the pubhsher is required lor all olher derivative works, mcludmg compilations and tianslalionsMectromc Storage. Permission of the pubhsher is required to störe clectromcally any material contamed m this Journal mclud-mg any article or pari of an article Contact Ihe pubhsher at the address indicated

L\iept (ΙΊ oulluu'd ubuM no pait «/ Ihn puMuatitm nun he icpiodiucd \toiul in a letnnal \\Mem 01 tianwnlled in amIm m m hy an) meam eleitioim muhanital pholoiop\int> l e t o i d i n t f o i ol/uinnc »illioin pi im H i i l t d i peimi s s««i a/ ilii'PublnlidNotice. No n">pon\ibility / s (js^unicd b) thc pubh^iu ι joi an) in/ui \ and/οι dumu^c lo pci Si>/?s 01 piopci i\ a\ a malte ι of pi ödm Λ Itabilil) nc^liyence 01 ollici M isc 01 // om ün\ »u ot opu ulion oj an\ ntclhod\, pi odin /s in s// iit lton\ 01 i(k'a\ ι onlaincd inthc muiei ml ha c in

Pubhshed monthly 01782166/97/11700 Pnnted m I h e Nethci lands

(% Fhe paper used m this publication meets thc requircmcnts of ANSI/NISO /19 48-1992 (Pcrmanence of Paper)

ELSEV1ER Journal of Pragmatics 27 (1997) 61-82

Interaction and causation:Causative constructions in modern Standard Dutch

Ane Verhageir1*, Suzanne Kemmerb

Centn foi Languaye and Communuatwn Uticcht UmvctMty Tiam, 10 3512 JK UtiahtThe Ncthcr/and·*

h Dcpaitment of Lmgunnt·; Rice Umvei vty 6100 Main Stieet Houston TX 77005 1892 USA

Rcceived January 1996 revised Version February 1996

Abstract

The Dutch verbs doen ('do') and taten ('let') categonze an event äs mvolving either director indirect causation, respectively The latter means that another force than the agent's is seenäs more immediately involved in bnnging about the effect, and is therefore especially suitedto indicate interactions betwcen humans i e mind-to mind causation The difference betweenthese verbs reflects the folk world view in which the mental world is seen äs separate fromthe physical, each having distinct causal properties We show how this explams the sharpdifference, observable m usage, in the preferences of both verbs for animate and inammateparlicipants

Another cultural cognitive model playmg a lole m the use of doen vs laten is the 'folkmodel of the mind', which governs our understandmg of mental processes such äs perceptionand behef Agamst the background ot this model, Speakers exploit the choice between the twoverbs for particular effects, e g to attnbute particular causal powers to certain referents, suchäs God or government authonties, or, combining verb choice with case markmg, to subtlyindicate different degrees of autonomy and affectedness of causees

The study demonstrates the intimate relation between cognitive models, pragmatic contex-tual factors, and lexical semantics

1. Introduction

This paper reports on one part of a larger project on causative constructions (andm fact, the hnguistic expresston of causation m general) In Kemmer and Verhagen(1994) we developed a general conceptual framework for handlmg cross-lmguisticmarkmg patterns m causatives Here we will apply that framework to an m-depth

* Corresponding aulhor Email anc verhagen@let ruu nl Phone +31 302538131 (office) +31 715174079(homc) Fax +31302536000

0378 2166/96/$15 00 Copyright © 1996 Elscvicr Science B V All nghts rescrvcdPH 90378 2166(96)00003 3

62 Λ Vcrluiiicn, S. Ki-mmer ι Jintrnal <if l*raf;maii<-s 27 ( I W 7 ) dl 82

study of a single language. Specifically, our focus is on ana ly t ic causa l ivc construc-lions in Dulch.

Modern Dutcli has t wo causativc vcrbs, doen and taten, which takc bare i n f i n i t i -val complemcnts, i.e. coinplemcnts laeking the i n l ' i n i l i v a l inarker le Cto'). Exampleswith bolh verbs are given in (l)-(2), and in (3)~(5), respectively ( u n d c r l i n i n g isadded only f'or c lari ty; it does not indicate emphasis):

( l ) De strafende zon doei de temperaluur oplopcn.tl ie shining sun does the ternperature rise'The bright sun makes the lempcrature risc.'

(2) De recessie doei_ de mensen verlangen naar helcre tijden.the reccssion does the peopie long lo bctter times'The recession makes peopie long l'or bctter times.'

(3) De as>ent Ijel he n passeren.the ol'ficer let them pass

(4) Z// Hei de af>ent haar rijhewijs zien.she iet the officer her drivcr's-liccnse see'She showed the officer her driver's license.'

(5) De sergeant liet ans door de modder kruipen.the sergeant lel us through the mud erawl'The sergeant had/made us erawl through the mud.'

Thus, it looks l ike there is at least some Var ia t ion (and perhaps competition) bctweentaten- and i/wu-construclions exprcssing causation.

Before we proceed wilh the analysis of t h i s Var iat ion, we need to introduee somebasic terminology. The causative verb (here taten or doen) expresses what we w i l lcall a 'causal predieate', i.e. some type ofcause; we w i l l be more spccific about thenature oi" the causal scmantics of these vcrbs in the course of this paper. The inl ' ini-tive in the eonstruelion expresses what wc w i l l eall the 'effeeted predieate': theprocess or state brought about by the causal predieate.

Effectecl predieates eome in two varieties: intransi t ive and t rans i t ive . It makes adifference in the overall scmantics of the eausal event, in general beeause theyinvolve two diffcrent eonfigurations of participants (äs deseribed in Kcmmcr andVerhagen, 1994). In the first type, there are two participants: a eauser and a eausec.The eauser is the init ial energy souree l'or the entirc composite eausal event - theentity eonstrued äs bringing it about.

The terminology is illustrated in (Γ): the eausal predieate is doet, the effeetedpredieate (in this case intransitive) is oplopen ('rise'). Eor brevity, we w i l l refer to acombinalion of a causal predieate and an intransit ive effeeted predieate äs an 'intran-sitive eausative'.

( l ') DejiiJ^ikndiLZiM- doel djnejnnjmiluuL oplopen (effeeted predieate (EP) intransitive]

'eauser' causee

Λ νΐ'ΐΐιαχαι, S. Kcnimei / .lourmil af PiiifiiMilicb 27 (IW7) (>Ι <S'2 63

The shining sun is the causcr (thc source of cnergy — in this cusc quite literally), andihe Icmpcraturc is thc causee, which we can provisionally deline äs the participantthat is thc immediate rccipicnt of thc cnergy supplicd by the causcr, and the one that'carrics out' the efl'ectcd prcdicate.

A causativc slructtirc with a transitive effccted prcdicate (zien, 'sce') is illustratcdin (4'):

(4') /(/ lict de agent haar rijbewijs zicn [BP transitive |

'causer"causec' 'at'fectee'

As in ( I ) , thcrc is a causcr and a causee, but there is also a third participant, whichwe call an 'affeclee': a participanl that is the final cndpoint of thc energy flow in thccntire causative evcnl.1 We call it 'affectec', bccause in protolypical cases, it isaffected in thc way that transitive objccts are affectcd participants. Again for brevity,we refcr to such struclurcs of causal predicatcs with transitive effccted predicatcs äs'transitive eausativcs'.

In Dutch, äs in niany languages, it is not at all unusual to find transitive eausativcswith no causee cxprcsscd, äs in (6):

(6) ////' het'ft ecn rolsloel Igten botiwen.he has a wheelchair let build'He has had a wheelchair built. '

This senlencc means that he caused sonicone, who is not identified any further, tobuild a wheelchair; thus the Interpretation of the causee, being left out, is highlyschemalic: because Ihe causee receives no independenl linguistic expression, itsInterpretation is exhanslecl by the Information provided by the effccted predicate,which evokcs ihe role of a 'buiklcr' (öl a wheelchair).

The definition of the causee givcn abovc is provisional because in fact the specificsemantics of thc causee is heavily depcndent on thc entire event structurc. It is pos-sible to gcnerali/.e over the two kinds of event structures, äs we demonstrated inKemrner and Verhagcn (1994), where the analytical framework is laid out in moredetail; for our purposes in this papcr, however, the definitions given are sufficicnt.

2. Some corpus data

The analysis we w i l l presenl in the following sections is largely basedon data l'rom an electronic corpus of (mainly written) Dutch, the Eindhoven

1 'lineigy I low" rclcrs not simply lo Iransmission of physical cnergy, but abstraft analogucs incorpo-ratmg nsymmetricul rclations in gcncral in thc cITccted prcdicate (c.g. 'scc' in (4)) are includcd (cf. Kem-mcr, 1994: 191--I92). The 'cnergy flow' in ihe causal prcdicate, also an abslract analoguc of physicall'orce, refers lo Ihe lypcs of forces laid out in Talmy's Force Dynamics framework, which will bc clabo-ratcd in thc course of our analysis.

64 A Verflogen, S Kemmei l loumal oj Piagmatin 27 (1997) 61-82

Corpus,2 which contains approximately 800,000 words, in 53,000 sentences; exam-ples used in this paper which are taken from the corpus are marked '{ec}'. Weextracted all sentences with the causal predicates doen and leiten, and determinedtheir distribution over the types of causative constructions described above. Somerelevant figures are given in Table l.

Tablc lMain distribution of types of causaüves

Total number of analytic causative construclions 855,- 439 intransitive EP (9 causceless),- 416 transitive EP,

- 272 (65%) causeelesb,- 144 (35%) with exphcit causee;

I.e percentage of three-participant causative constructions 17%

- Total number of causative leiten 686 (444 with exphcit causee)- Total number of causaüve doen, 169 (130 with exphcit causee)

First, notice that there are considerably more cases of loten than of doen. Further,the intransitive causative structures slightly outnumber the transitive ones. Butwithin the transitive sei, the majority are causeeless causatives like (6) - conse-quently, the percentage of the total with a fully elaborated three-participant structureis relatively low, in fact just 17%.3 As will beconie clear in the course of the analy-sis, this 17% is the locus of the most complex patterning.

We are particularly interested in the conceptual relation between causer andcausee and in its linguistic expression. Therefore, we separated the cases with anexplicit causee for further examination: almost all (430 out of 439) intransitivecausatives, and the minority (144 out of 416) of transitive ones (thus N = 430 + 144= 574). In view of our cross-linguistic findings äs reported in Kemmer and Verhagen(1994), the first feature worth considering is the distribution of animacy over causerand causee; 'animacy' äs we counted it includes not only humans, but also humaninstitutional entities, like the government. The results are given in Table 2.

There are two main patterns that emerge from these counts, i.e. two patterns ofheavy skewings in the frequency of different animacy categories within the subsetsfor leiten and doen. The first pattern concerns the causers. Notices that with loten, theoverwhelming majority (99%) of cases are animate; with doen, on the other hand,there is a preference for manimate causers - albeit not nearly äs strong äs the reversepreference with leiten.

2 The version of the corpus we used is the one available Irom the Frec Umversity dt Amsterdam. It isdescribed in Uit den Boogaart (1975) and in Rcnkema (1981)3 This tendency towards, causeeless causatives was already noted by Dik (1980 81) In fact, our con-clusion is somewhat süonger than Dik's, the reason being thal he only lookcd at a subset (of 594) of thetoe«-cases in the Eindhoven Corpus, i e. those he considercd causative (excluding those hc consideredpermissive)

A Vcihatten S Kemmei l louinal of Piagmatn\ 27 (1997) 61-82 65

Tablc 2Distribution öl ammacy öl causcr dnd cdusce

Causer diiunatecausee animatecausce manimalc

Causer mammatecausce animdtecausce inanimate

Laten (444)

99%

1%

49%51%

17%83%

Doen (ΠΟ)

42%

58%

21%79%

58%42%

Moreover, in those cases where we do have an animate causer with doen, there isa large skewing m numbers for the animacy of the causee· The causees that are man-imate are almost 4 times äs frequent. An example of this Utter kind of causative is(7)

(7) We zullen de ι eorgam vatie gefaseerd doen plaatsvindenwe shall the reorganization m-phases do take-place'We shall have the reorganization take place m stages.'

We shall return to this specific kind of example, fairly typical for a particular kmd ofdiscourse, in the course of the analysis.

Within the same subset of the data compnsed by causatives-with-causees, there isa second set of patterns to be noticed - agam, one that proved to be highly relevantcross-hnguistically in our previous study. This concerns the causees, which mayeither be preposition-marked or have zero markmg Consider the data m Table 3.

Tdblc 3Case markmg of causees

Always zcro when - causal prcdicale is doin OR- EP is intransitive

Possibly prepositional when - causal predicatc is loten, AND- EP is transitive(cascs satisfymg both conditions 118, 14% of all causatives)

- zero 56 mstances (9 8% of all causees)- dooi (agcntive/instrumental) 55 mstdnces (9 6% of all causees)- aan (dalive) 7 mstances (l 2% of all causees)

There is no prepositional markmg of the causee with intransitive causatives, andnone with doen either (regardless of transitwity) But m transitive causatives withtaten - i.e. within a subset of the already relatively small set of three-participantstructures (cf Table 1), we do find difterent options for markmg the causee; thissubset contams 118 mstances. Here the Variation is considerable. We find three pos-sibihties· there is no preposition, or the preposition is door, which mdicates means

66 A Vuhaj>cn S Kcmmei /Journal t)/ Piaqmatii s 27 (1997) 61-82

and passive agent (cf Cornehs, 1994), or the preposition is aan, which has dativefunctions As the table shows, zero marking (no preposition) and doot have aboutthe same number of mstances, while there are relatively few mstances with aan (lexamples m total)

Sentences (8) and (9) exemplify a dooi -marked and an aan-mdrked causee,respectively

(8) Zij wilden Woody daaina door een echtpaai laten adopteren <ec)they wanted Woody thereafter through a marned-couple let adopt'They wanted to have Woody adopted by a marned couple '

(9) Laut dit aan niemand lezen <ec)let this to nobody read'Don't let anybody read this '

By now we have two sets of patterns the skewmg m frequency of animate andinammate causees with doen and laten, and the distnbution of case on the causeeWe will now provide an analysis that accounts for both sets of patterns, which arenot obviously related, withm one general conceptual framework

3. Doen vs. laten: Direct and indirect causation

31 Laten Permission and induect causation

Let us begin by considenng the semantics of doen and laten, in order to show howthis bears on the analytical problem 4 Each oi these verbs occurs not only mcausatives, but also in simple clauses without an mfimtival complement With laten,it is clear that the nolion of 'allowing', i e permission or enablement, is central to itscharactenzation m such simple clauses, (10) is the simplest kind of example

(10) Ik laat IQU de keusI let you the choiceΊ leave the choice to you '

Here, the Initiator has some power to either grant or prevent something, and grantsit, the sentence may be paraphrased äs "I do not prevent you from making thechoice" In (11), we have a locational complement, the sentence says that sheallowed the cat to change its location, (l la), or eise to remain m its location, (l Ib)

4 This analysis is also useful to illummate the coherence m Ihc uses öl both doen and taten äs thcse aregiven m dictionanes (esp Van Dale Groot Woordenboek der Nederlandse Taal 1992P and the com-prehcnsive Woordenboek der Nederlandsche Taal (WNT)) A survcy of the Information in thcf>e refercnce works showcd that what we say about the meaning of doen and taten has general vahdity See alsonotc 9

Λ Veihagen, S Kemmei l louinal oj Piabmatte s 27 (1997) 61-82 67

(11) Zij liet de kat m het huis.she let the cat in the house(a) 'She let the cat into the house.'(b) 'She left the cat in the house.'

The ambiguity comes out m the English translations: the first reading can be para-phrased äs "She allowed the cat to go into the house", i.e. to change its state/loca-tion, whereas the second readmg means "She allowed the cat to stay in the house",i.e. to remain m its state/location. Examples (12) and (13) show something similar,but now with a pure state, indicated by an adjective, rather than a location; (12) hasthe Interpretation of a change of state (cf. (Ha) above), while (13) has the Interpre-tation of something remaining in the state it is in (cf. (llb)):

(12) We beten de zeehond vri].we let the seal free'We let/set the seal free.' (i.e.: we granted it the state 'free')5

(13) Schep deze saus over de asperges (maar lagt de kappen vnj).pour this sauce over the asparagus (but let the heads free)'Pour this sauce over the asparagus (but leave the tips uncovered).'

Turmng now to combinations of leiten with infinitives, we claim that it is bestcharacterized äs marking indirect causation, which may be viewed äs an extensionof the notion 'permission/enablement'. Indirect causation can be defined äs a Situa-tion that is conceptualized in such a way that it is recognized that some other forcebesides the Initiator is the most immediate source of energy in the effected event.6

That is always true in cases of permission and enablement, but the reverse does notnecessarily hold - i.e. 'indirect causation' subsumes permission and enablement, butalso other indirect causal events. Let us make this more concrete by discussing someexamples. First of all, consider (14):

^ Whcn adjacenl, ihc combmalion of vn/ and laten may in ccilam circumstances be viewed äs a com-positc verb (vn/laten), vu with a parlicular combmation of phonologtcal and scmantic fealures. when ilhas the stress pallern of a singlc word (wilh primary stress on Ihe first elemcnt, v;;/-), it only has Ihcmeaning Ίο sei iree', (cf the plural form de vn/gclotenen, lit 'Ihc hberatcd ones', i e frced slaves)Naturally, the fact that the composite form is in some sense a unit (with a combmation of phonologicaland scmantic tcaturcs lhal is not dcnvable compositionally), docs not mean that it is no longei analyz-ablc (see Langacker, 199Γ 261-288, lor gcneral discussion)6 Turnei (1987, esp Π9-183) convinungly shows that many instances of causation cannot be charac-tei i/cd adequatcly m Standard conccptions of causahty, such äs necessary and sufficient conditions. Hcthen argucs lor a model that he calls 'causation äs progcneration' a cause is melaphoncally conceptual-ized äs, a parent pioducmg the resull äs ollspnng. One ot Ihe advantages of this model is preciscly thatU allows Speakers lo conccivc of all kinds of relations äs causal withoul bcmg committed to a spccifica-tion of all possibly relevant intcimediary factois (äs wc conceive of olfspnng äs produced by paients,without our bcmg ablc to specily all necessary and sulficient conditions tor such a proccss) In the samevom, indirect causation allows Speakers lo presenl a certain result äs causally dependent on some agent,while leaving room for othei factois than just those mentioned in the sentencc.

68 A. Verhagen, S. Kemmer l Journal of Pragmatics 27 (1997) 61-82

(14) Hij haalde de stop eruit en lief he t badwate r weglopen.he took the plug out and let the bath-water run-away'He took out the plug and let the bathwater flow off.'

In the event äs presented in (14), the bathwater flowed out by virtue of some moredirect source of energy than the motion of pulling the plug (i.e. gravity, for educated20th-century Speakers). This is essentially also Talmy's (1976, 1988) force dynamicaccount of English let, and it is clear, of course, that enablement/permission providesa good characterization for (14), too. But the more abstract characterization äs indi-rect causation fits it äs well, for example since the Initiator may be consideredresponsible for the consequences. The reason that we claim that Dutch laten has tobe characterized in terms of the more schematic sense of indirect causation, ratherthan the specific sense of enablement/permission, is that in the Dutch causative con-struction, laten always has the sense of indirect causation, but not always the enable-ment/permission sense. Consider example (5), repeated below:

(5) De Sergeant liet ans door de modder kruipen.the Sergeant let us through the mud crawlThe sergeant had/made us crawl through the mud.'

The most likely reading of (5) is that the sergeant actually has exerted someauthoritative force (an Order' of some kind) on us, so that we had to do it. Thus (5)does not indicate permission but ralher coercive causation. Still, categorization äsindirect causation is justified because we had to move under our own power: thesergeant has no direct control over our bodily movements, and still has to, in someimportant sense, 'communicate' in order to get his order carried out. Thus, the Otherforce' that is recognized äs most immediate source of energy in the event is not nec-essarily restricted to some immanent, inherent tendency in the causee or in the world- it may very well also be another force that is induced by the causer himself. Thisis true not only with animate causees, but also with inanimate ones, äs is illustratedby (15):

(15) [Er wordt in de laboratoria doorlopend geexperimenteerd om de kunstvezels tevervolmaken.J Om ze nog meer te laten lijken op echt haar, (ec)'[There are continuously experiments going on in order to perfect the syntheticfibers.] To make (lit.: let) them look still more like real hair.'

There is no inherent tendency in the synthetic fibers to look like real hair (on thecontrary, in fact), but the event is still conceptualized in such a way that the experi-menters themselves are not the immediate cause of the hair-like appearance of thefibers; probably it is some physical-chemical process induced by them (whose nat-ural result is that the synthetic material looks more like real hair).7

7 This example shows one interesling difference between Dutch laten and English let: laten allows forIcss autonomy of the causee than English let. We leave an analysis of the differences for later work.

A. Verhagen, S. Kemmer l Journal of Pragmatics 27 (1997) 61-82 69

In fact, what we find with laten is a continuum of indirectness. At one end(enablement/permission), there are cases with relatively great autonomy of thecausee and some inherent tendency for it to carry out the effected predicate; anexample would be (3), under the (most likely) reading that it was the causee's wishto pass:

(3) De agent liet he n passeren.the officer let them pass

At the other end (coercive-causative) are cases in which those factors are muchmore reduced and the causee would not normally carry out the effected predicate inthe absence of an external force; the example already given is (5). There are also in-between-cases, like (4):

(4) Zij liet de agent haar rijbewijs zien.she let the officer her driver's-license see'She showed the officer her driver's license.'

Here, the use of laten might be motivated on the grounds of a request by the officer(with or without a reference to his authority), which would make it look like per-mission; but it is also motivated in the Situation in which she has asked the officerto look at the driver's license and he complies with her request (causative), for theseeing still involves some processes on the part of the officer which are not underdirect control of the Initiator. Thus, in cases like (4) and many others, it does notmake much sense to try to decide between a permissive and a causative reading: theuse of laten, categorizing an event äs involving indirect causation, in itself simplyleaves this undecided. Note that the same possibilities (showing on request, vs.showing not on request) are in fact present in the English translation of (4), and thatthis does not lead one to argue for ambiguity of the verb show*

Dutch laten can thus be seen äs having a fairly schematic meaning. It may refer toenablement, or its corresponding concept in the sociophysical realm, permission: theInitiator has the power to prevent the occurrence of the effected predicate, and doesnot do so, thereby leaving a second force (natural or human) -to come into play ineffecting the event. In these cases the role of the Initiator is a relatively passive one.Laten is, however, licensed in a much wider array of contexts, including instances ofactive instigation sometimes amounting to outright compulsion. What these contextshave in common is the conceptualizer's recognition of an intermediary force thatmost directly brings about the effected predicate.

All in all, we claim to have established that the use of laten in combination withan infinitival complement uniformly marks the causality in the event äs being indi-

8 Thus, in our view it is not insightful to scc the difference between permission and causation äs amatter of lexical ambiguity (äs in, o.g., Dik, 1980). Rathcr, it musl be a matter of the construal ofInterpretation at (minimally) the clause level, taking all other factors into account, and furthermore amatter of degree.

70 A. Verhaken, S. Kemmer l .Journal a/ Pragmatik!* 27 (1997) 61-82

rect, and that this use is closely connected with the enablement/permission that ismarked by leiten in other uses.

3.2. Doen: Producing a result and direct causation

Turning now lo doen, a consideration of its use in simple clauses makes it clearthat a central aspect of its characterization must be the exertion of energy by the ini-tiator of the event depicted; (16) and (17) provide examples with animate and inan-imate subjects, respectively:

(16) Vandaag doe ik examen.today do I exam'Today, I do/take my exam.'

(17) Dal doet pijn l doet mij verdriet.that does pain / does me sorrow'That hurts / grieves me.'

Sentence (16) means that I perform an activity, putting energy into producing a cer-tain result (e.g., answers to the exam questions). The case with the inanimate subjectin (17) means 'produce an effect', which comes very close to causation, of course.In both cases, there is no intervening energy source 'downstream' (cf. Langacker,1991: 217) from the Initiator: if the energy is put in, the effect is the inevitableresult. This sense of directness is also present in combinations of doen with infini-tives, i.e. in causatives. Consider example (1), repeated here for convenience:

(1) De st ratende zon doet de temperatuur oplopen.the shining sun does the temperature riseThe bright sun makes the temperalure rise.'

The rising of the temperature is the inevitable consequence of the sun shining. Thecharacterization of causative doen is thus definitely distinct from that of taten. To besure, this does not mean that for all conceivable or actual contexts, only one of theseverbs will be allowed. Some situations will allow a Speaker to naturally conceptual-ize a specific event äs involving direct causation äs well äs indirect causation.Despite this possibility, though, the choice of one verb rather than another will oftenresult in a specific effect, due to the semantic difference. We will discuss examplesof this phenomenon in Section 4.2.

4. Cultural cognitive models of causation

4.1. Models of causation type s

The characterizations of laten and doen allow for a very nice mapping onto someof the distinctions between types of causation proposed by Talmy (1976, 1988), inparticular äs those distinctions are schematized by Croft (1991):

A Veihagen, S Kemmei l touinal of PiagmatK v 27 (1997) 61-82 71

INITIATOR ENDPOINT

'MENTAL'

THYSICAL'Physical

Fig. 1. A model of causalion types (Cioil 1991 167; bascd on Talmy 1976)

Causative events are distinguished along two dimensions. The first is the dis-tinction between Initiator and endpoint of the causal change. Notice that the modelis highly abstract and schematic, generalizing over both simple and more compli-cated causative events. The claim is that all of them involve at least an initiatingelement, and an endpoint; the endpoint may correspond to the (state of the) causeein intransitive causatives (of the type 'He let/made the baby cry'), or to theaffectee in transitive causatives (of the type 'She let/had him bake a cake'), inwhich case the causee has a more intermediary role to play (Kemmer and Verträ-gen, 1994).

The other dimension is essentially the distinction between animate and inanimate.An obviously important aspect of this model of causation types is the very markedasymmetry between entities with a mental dimension (animates) vs. those that aremerely physical. Animates can only act on animates via the intervening physicalworld, i.e. the model implies that one cannot reach into another person's mind anddirectly cause him or her to do, feel, or think something. Physical entities are takento act directly on other things; hence the straighl arrows in the diagram in Fig. l, vs.the very bent arrow for mental-on-mental causation, and the slightly bent one formental-on-physical.

This model immediately provides a basic understanding of the animacy skewingobservable in Table 2. If loten indicates indirect causation, one would expect it to beparticularly suited for indicating inducive (mental-on-mental) causation, and so tohave a relatively high frequency of animate causers, which is just what the tableshows.

Doen, on the other hand, ought to occur more often with /«animate causers, whichit does. Moreover, the place where doen is least expected to be found is with ani-

72 A. Verträgen, S. Kemmer l Journal of Pragmatics 27 (1997) 61-82

mates acting on animates - and this indeed corresponds to the cell where the minor-ity of animate causees within the minority of animate causers is to be found.9

In terms of Lakoff's (1987) notion of 'Idealized Cognitive Models' that are orga-nized in terms of prototypical centers and radial extensions, we may say that severalICMs of causation are involved here, which map onto two Schemas of direct andindirect causation, expressed by doen and laten, in a fairly straightforward manner.Inducive causation clearly is a prototypical case of indirect causation, and thus it isexpressed by means of laten quite generally. Physical and affective causation areclear instances of direct causation, thus they are generally expressed by means ofdoen. Volitional causation is, in terms of categorization äs direct or indirect, the mostcomplex, i.e. neither prototypically direct nor prototypically indirect; thus it comesäs no particular surprise that quite a number of examples of both doen and laten areto be found in this subclass: volitionality is in itself not a heavily weighted factor inthe categorization of the event äs either direct or indirect, and thus other factors willmore often be decisive here than in other subclasses. ßut the same Other factors'may sometimes (less frequently than with volitional causation) also Outweigh' thecausation type in the other subclasses — in particular, in inducive causation. It istherefore interesting to see if and how our analysis can account for the data in a morefine-grained way; this is what we will do in the next section, where we will pay spe-cial attention to some direct contrasts between doen and laten, and to uses of theseverbs in some less prototypical cases äs these are defined by our model.

4.2. The folk model of the mind

What we will do now is to further enrich our analysis by invoking some details ofthe 'folk model of the mind', how the mind relates to the physical world, and howevents in the mind may be caused. In Fig. 2, some relevant aspects are presented ofthe folk model of the mind prevalent in (at least large parts of) Western Culture, äsit has been described by D'Andrade (1987).

According to this model, there is only one mental state that is conceived of äscaused directly by the outside world, viz. perception. That is, (despite modern theo-ries of vision, so to speak) we think that we see what we see because it is there,within our visual field, in the outside world; and we cannot avoid perceiving it if itis there - perception is not controllable.

On the other hand, a mental state like my believing something is conceived of äscontrollable, so it is not possible for something in the outside world to immediately

9 We want to lhank an anonymous rcvicwer for drawing our attcnlion lo the Dutch-English translationdictionary Van Dale (Utrecht and Antwerpen 1986). Under laten this gives 'drop something' for ietslaten vallen (lit.: Ίο let something fall', p. 708), and under doen it gives 'a stone made himstumble/tripped him up' for een steen deed hem vallen (lit.: 'a stone did him fall', p. 296; note that theagent in this example is inanimate). This is indeed anothcr nice Illustration of our general point: In thesecond case, the imagc of the event is such that the presencc of the stone ilself produces the result (itblocks the normal proccss of walking), while in the first case there is no such immcdiatc connectionbetween what the agent does (no longer holding something, thereby 'allowing' gravity to excrt its force)and the result.

A Veihagen.S Kemmei l Jouinal of Piagmatics 27 (1997) 61-82 73

Perception Belief Feelmgs Desires Intentions Resolutions

cause outside causc insidemmd ramd

not controllable usuallycontrollable

cause msideand outsidemmd

usually notcontrollable

cause inside cause insideand outside mmdmmd

not controllable controls itself

cause insidemmd

control of control

Fig 2 Aspects of the folk model of the mmd (D'Andrade, 1987 117).

cause me to believe a certain proposition — at least so we think. We will return toother mental states below, but let us first consider some consequences of these rela-tively simple observations.

If we have an event with a mental endpoint and an outside causer, and it is cate-gorized äs direct causation by means of doen, then it is naturally taken äs involvinga case of perception. But if we have such an event categorized äs mdirect causationby means of leiten, then it is of course naturally interpreted äs a case of inducive cau-sation, with the causee äs an intermediary in the process. Now consider the differ-ence between (18) and (19):

(18) De psychiater deed mij aan mijn moeder denken.the psychiatrist did me at my mother think'The psychiatrist made me think of (reminded me of) my mother.'

(19) De psychiater lief mij aan mijn moeder denken.the psychiatrist let me at my mother think'The psychiatrist had/made me think of/about my mother.'

The first of these, with doen, does indeed describe a case of perception. The psychi-atrist is. in fact not involved in the event äs an intentional animate being - he neednot be present or even alive for the sentence to be usable (adequately). Rather, it issomething observable in the way the psychiatrist looks or behaves that causes me toperceive a certain similarity. Example (19), on the other hand, is completely differ-ent. Here, the psychiatrist and l are communicating. He/she intentionally teils mesomething, and I decide to follow his/her advice; my thinking of my mother here isalso deliberate on my pari.10 So while (18) does not describe interaction betweenhumans, (19) does, and it is evident why this should be the case, given the meaningof doen and taten and the folk model of the mind.

These observations explain the occurrence of a few of the sentences with causerscounted äs animate (to be precise: 3, i.e. 5.5% of animate causers with doen). Thereason for such occurrences is that this count was done independently of other fac-

10 As an anonymous revicwer pomted out, such dehbcrale thinking äs m (19) allows for the use of thepreposition ovei ('about' - denken over means Ίο reflect on', 'to contemplate'), while replacmg ovei foraan in (18) leads lo an inappropnale sentence

74 A. Verhagen, S. Kemmer l Journal of Praf-maticx 27 (1997) 61-82

tors than descriptions of the referents themselves, i.e. independently of the construedanimacy of the Initiator. So noun phrases like De psychiater were always counted äsanimate, based on their inherent semantics. What we see here is that the use of doenmay be precisely the decisive factor in determining that the feature 'animale',although available from the lexical meaning of the head noun, actually plays no partin its role äs causer in this particular event.

Recall now example (2), repeated here for convenience.

(2) De recessie doet de mensen verlangen naar betere tijäen. (ec)the recession does the people long towards better times'The recession makes people long for better times.'

This sentence describes a feeling or a desire (perhaps both). Given that according tothe folk model of the mind we do not think of such mental states äs controllable, itis again understandable that we find doen here. The recession is not another mind,interacting with the people, and the desire for better times is also depicted äs beyondthe people's control: it is not viewed äs the result of a deliberate decision, it just isthere äs a consequence of the perception of the recession.

In Section 4.1, we pointed out how our analysis accounts for the generaltendency in the skewing between animacy and type of causation; here we have sofar added some considerations deriving from the folk model of the mind, whichjustify why certain kinds of events are categorized äs direct or indirect causation.A problem still to be addressed, however, is the fact that the percentage of doen-cases with an animate causer is, although a minority, still considerable (42% inTable 2); most specifically, we still face the question why the cell of doen withboth an animate causer and an animate causee is not completely empty. Theconceptual question behind this is: What does it mean for a causal relationshipto involve a mental Initiator and a mental endpoint, and still be categorized äsdirect?

In order to answer this question, we draw attention to the specific nature of com-munication, äs, in a way, implied by the folk model of the mind. Inherently, com-munication is indirect. As pointed out above, no human mind can directly cause achange in another human mind. Categorizing such events äs indirect causation withlaten recognizes this fealure, and thus recognizes the separate, partly independentrole of the target of communication in bringing aboul the intended result. Categoriz-ing such an event äs direct causation, we may now say, therefore makes the heareror reader focus exclusively on the part of the action of the causer that is completelywithin its own control, or from another perspective, it presents the entire event äsbeyond the control of the causee.

For example, God can be conceived of äs capable of directly causing anything,including a change in a person's mind (cf. the conceived nature of the phenomenonof conversion). Consider example (20),

(20) Pater Germano geeft haar de raad, Jezus te smeken, haar de gewonefather Germano gives her the advice, Jesus to beg, her the ordinary

A Vuha^n S Kemmci / louinal of Piaqmatia, 27 (1997) 61-82 75

weg te doen bewandelenroad to do walk-on'Father Germano counsels her to beg Jesus to make her walk m the usualpath '

The way this is to be understood is that the 'her' is to beg Jesus, not to commumtatewith her, but to directly make her do 'the nght thing', ovemdmg her own vohtion ifnecessary This type, with God äs causer, accounts for another three cases (5 5%) ofthe c/oe/7-causatives with mmd-to-mmd-causality

Another subset (of the same size) is the followmg Smce it is not normally possi-ble for one person to immediately create a belief in another person (a behef bemgunder that person's own control), it should only be possible to readily use this typeof expression for non-actual events And we do in fact find some combinations ofdoen with geloven (meanmg 'to believe'), but none of them is assertive Considerexample (21)

(21) wat vommige tethmti oni ook willen doen geloven (ec)what some techmcians us also want-to do beheve

'whatever certain techmcians want to make us beheve ' (i e , they cannot makeus beheve it)

The use of the complete phrase ("Such-and-such is the case, whatever they want tomake us beheve") m fact imphes that they will not make us beheve it, which is justwhat our approach imphes Thus, the fact that combinations of doen and geloven('beheve') occur m the corpus m piecisely this wayn m fact confirms our approach,rather than that it contradicts it

Some more mtncate, but also very illuminatmg, examples are to be found in (22)and (23)

(22) Met een zucht deed hij de buitenwereld weten dat het kleine veitiekwith a sigh did he the outside-world know that the small roombezet was (ec)occupied was'With a sigh he made it known to the outside world that the small room wasoccupied ' [i e , humorously "He locked the door of the bathroom"]

(23) Gaaine wil ik u doen weten dat ik geen enkelegladly want I you do know, that I no smgleverantwooi delijkheid kan netnen vooi de nu mtgevoeideresponsibihty can take for the now executed

1 ' The other two mstances of this type m the torpus are hypothetitals (oi the type They would wantto mdkc us beheve ) agam evokmg the opposite of actual An independent search for doen + gelovenm Dutth and Flemish weekly maga/mes (Nathahe Lans p c ) strongly Supports this generaluation allcases found werc non actual (of Ihe type descnbed here with They wanted to make someonc beheveäs the süongest tase)

76 A Verhagen, S Kemmer l Jouinal of Pragmahcs 27(1997) 61-82

werkzaamheden. (ec)Operation sΊ want to teil you that I cannot take any responsibility for the operations thathave been executed now.'

Unlike previous examples, these neither involve God, nor denial of direct mentalcontact, but they still conform to the general analysis. The interesting point of (22)is that it evokes, in a humorous, ironical way, both the communicative and the non-communicative aspects of the Situation. The physical act referred to is the sliding ofthe latch of the bathroom door, with the effect that there is a signal on the outsideindicating Occupied'. The effected predicate is 'know' — something mental; but thecausee is 'the outside world' — not very human, and not very specific; and the causa!relation is categorized äs direct, by means of doen. Mainly because of this verb, thenon-communicative aspect is foregrounded: the causer changes something in thephysical world, the result of which might be interpreted, by whoever might comealong äs This bathroom is occupied' - we understand the sentence to mean pre-cisely that no actual person is construing this Interpretation: doen implies that thereis no real communication.

Such examples provide a clear Illustration of a more general point. They show thatit is not really possible to set up selectional restrictions in any strict way; for exam-ple, even though it might look plausible at the Start, we cannot stipulate a rule to theeffect that mental effected predicates select leiten. What is actually going on is thateach lexical and grammatical signal chosen by the speaker/writer sets up a constraintfor the hearer's/reader's Interpretation; the latter must, äs a whole, maximally satisfythe set of constraints presented in the utterance, but it is clearly incorrect to say thata given element absolutely constrains the occurrence of another element. In fact, theearlier examples with God äs causer and non-affirmalive cases of doen geloven('make believe'), illustrate the same point.

In (23), the effect of doen for the Overall Interpretation of the sentence is that theinfluence of the causer is maximized. The author seems to want to guarantee thearrival of his message with maximal certainty, äs though it were a physical conse-quence of the way the world is. Categorizing it äs indirect by means of loten wouldmake this change in knowledge-state partly dependent on the causee. Put differently:by using doen the author focuses attention on an action that is completely within hisown control, and suggests that this is sufficient for producing the desired effect. Webelieve that this is what underlies the fact that in the Eindhoven Corpus the subcor-pus of political language - with government authorities äs causers - is the only onein which doen outnumbers laten.

In fact, we believe that these effects of presenting an act äs non-communicativeprovide the general motivation for cases of doen with inanimate causees, where theuse of laten would suggest an unmentioned intermediary person co-responsible forthe result. Consider the following examples.

(24) De bezetters hebben inmiddels een dreigende verklaring doenthe occupiers have meanwhile a threatening declaration make

A. Verhagen, S. Kemmer l Journal afPragmatics 27 (1997) 61-82 77

uitgaan. (ec)go-out'Meanwhile, the occupiers have sent out a threatening Statement.'

(25) De regering stell zieh voor deze herstructurering gefaseerd te doenthe government envisions [REFL] this restructuring in-phases to makeplaatsvinden. (ec)take-place'The government intends to have this reorganization take place in stages.'

Since the effected predicates are intransitive, and the causees inanimate, we may saythat these cases do not represent inducive, but rather volitional causation, and thatthis allows for the use of doen. However, that does not yet explain the fact that thisuse also has a particular semantic/pragmatic effect, in comparison with laten - whichwould in both cases have been possible too. By using doen, the reporter who wrote(24) explicitly categorizes this particular event (of the occupiers producing a State-ment) äs nothing more than sending out a piece of paper into the world, i.e. a non-communicative event; and s/he thereby also categorizes the occupiers äs not com-municating, or perhaps better, äs impossible to communicate with.

In the event referred to in (25), it is obviously true that people other than thoseconstituting the government itself will have to perform certain actions in order forthe reorganization to take place. However, by using doen, the author reduces the pos-sible influence of these intermediaries to practically zero, äs if the government'swish will suffice for the reorganization to take place in this particular way. Theresult is presented äs something that inevitably follows 'if the government says so'.These kinds of formulations therefore sound more authoritarian than if laten wereused - even when no human causee is explicitly mentioned - because the latterleaves more room for the inference that other forces than those mentioned, in partic-ular human beings with other intentions, might possibly change the outcome.

Summing up so far, we have shown that the idea that a causative event is catego-rized äs direct by doen and äs indirect by laten, in fact provides a good Instrumentfor explaining not only general distributional phenomena, but also intricate semanticand pragmatic aspects of sentences that might at first glance appear to be exceptionsto these gross patterns of distribution.

S. The marking of causees

As we mentioned in Section 2, explicit case marking of causees in modern Dutchis possible only in a fairly small subset of the instances of causative constructions inour corpus. We furthermore observed that it in fact only occurs with laten, and notwith doen (cf. Table 3). The explanation for this distribution is, we claim, simplythat it is only laten, because of its semantics of indirect causation, that leaves room,äs it were, for a ränge of possibilities for the contribution of the causee to the entireevent. It is only indirect causation that allows a construal of the causee äs a moredirect source of the force producing the effect than the causer. Especially in indirect

78 A Veihagcn S Ktmmcr / fouinal oj Piagmattct 27 (1997) 61-82

transitive causatives, containmg a separate 'affectee' that is the most affected participant, causees can therefore be construed with different degrees of autonomy andaffectedness At the same üme, the meamng of indirect causation is not specific ästo the actual degree of autonomy or affectedness The existence of a ränge of possi-bihties, we claim, motivates the potential for differences m exphcit markmg on thecausee m indirect transitive causatives We will now illustrate how the particularmarkmgs that occur are precisely those which mvolve specific degrees of autonomyand affectedness of causees

In Kemmer and Verhagen (1994), we argued that dative and instrumental markedcausees cross-lmguistically tend to mdicate greater autonomy and less affectednessthan accusative or zero marked causees, and that this m turn instantiates a more gen-eral pattern m which recipients, Instruments, and the hke, are more penpheral m, andless affected by, an event than objects are Dutch is no exception to this pattern Con-sider (26)-(28) 12

(26) Hij het de brief aan icdereen lezenhe let the letter to everybody read'He let everybody read the letter'

(27) Hij het de brief door temand lezenhe let the letter by someone read'He had the letter read by someone'

(28) Hij het haar de brief lezenhe let her the letter read'He let/had her read the letter'

Because of the dative markmg in (26), the causee is categonzed äs a recipient Thus,everybody is reading the letter for its contents The causee has a fair amount ofautonomy in the event, it clearly has a contnbution to make of its own, and the sen-tence thus tends towards a permissive reading "He let the letter read 'to' everybodywho wanted to"

In (27), the causee is categonzed äs an Instrument One possible reading is that thepurpose is to get the letter corrected by having someone go over it Another readingis that the letter is bemg read aloud by the causee, in order to get the message tosome other, unmentioned audience Thus, a natural continuation of (27) would besomething of the type ' and this person did a fine job', while the analogous 'and they did a fine job' would be highly mappropnate following (26) In any case,the reading by the causee is not for content, and the causee is not very much affectedby the event The sentence thus tends to be interpreted not äs permissive, but äs

12 Wc have taken these examples from Comnc (1976) and adaptcd them The adaptations conccrn thedescription of the causee m Order to bring it in linc with semantic effects of the prepositions Comnehad the pronoun haar ( her ) in all examples and furthcrmorc at the same position äs well We also wantto repeat here tht observation by Dik (1980) vi/ that contrary lo what Comne s presentalion imphcsthe pattern of a three way choitc with the same effccted predicatc is the exception rather than the ruieAs far äs we can see the verb lezen ( to read ) is even unique in this respeu

A. Vertagen, S. Kemmer l Journal of Pragmatics 27 (1997) 61-82 79

'just' causative (the causee being no more than an Instrument, the question of bis/herwishes with respect to the event does not really arise).

Sentence (28), finally, precisely allows for the highest degree of affectedness of thecausee. The default Interpretation is that the reading of the letter (i.e. its contents)really affects the causee; a continuation of the type '... and she did a fine Job' is notreally appropriate. This is the only type of causative sentence with laten that allows theInterpretation that the causee reads the letter more or less involuntarily, or under somekind of order. This sentence, too, is interpreted causatively rather than permissively.

Thus, it is specifically dative-marking with aan, äs opposed to instrumental andzero marking, that reinforces a permissive reading of laten. This comes out clearly ina pair like the following:

(29) De sergeant heeft ans het nieuwe kanon laten zienthe sergeant has us the new gun let see'The sergeant showed us the new gun.'

(30) De sergeant heeft het nieuwe kanon aan ans laten zienThe sergeant has the new gun to us let see'The sergeant showed the new gun to us.'

While (29) is neutral on the question of whether we wanted to see the new gun ornot, (30) favours the Interpretation that we did; for example, we may have requestedto see the new gun, and the sergeant was willing to please us - i.e. laten has a per-missive Interpretation here. In the event described by (29), in contrast, it may verywell be that we actually did not want to see the new gun at all (note that it is unnat-ural to have (30) preceded by an adverbial phrase like Zeer tegen onze zin, 'Verymuch against our wishes', while this is no problem in (29)).

The verb lezen, used in the examples (26)-(28), is the only one that allows allthree possible markings of the causee. However, the pattern of interpretations illus-trated by means of those examples is clearly observable in the corpus data äs well.Consider the following examples.

(31) Hij wilde het op Schiphol aan een collega laten zien. (ec)he wanted it on Schiphol to a colleague let see'He wanted to show il to a colleague, at Schiphol.'

(32) Hij wilde juist taten merken aan de mensen dat ze niet zo siechthe wanted actually let notice to the people that they not so badwaren als ze zelf dachten.were äs they seif thought'He actually wanted to let people realize that they were not äs bad äs theythemselves thought they were.'

Sentence (31), with aan, suggests that the colleague will be happy to see whatever itrefers to (he probably wants it), so he is not just a passive receiver. And in (32), withthe subject referring to Jesus, the people are no passive receivers either, but ratherbeneficiaries, who are understood to enjoy the message.

80 A. Vertagen, S. Kemmer l Journal of Pragmatics 27 (1997) 61-82

On the other hand, sentences like (33) and (34), with the same or similar effectedpredicates but without ααη-marking of the causee, have a more obviously causativeInterpretation, and indicate a more passive role of the causees.

(33) Charlie Greene liet de ruim drieduizend toeschouwers zien datCharlie Greene let the over three-thousand spectators see thathij ook op de sintels in alle opzichten uit de voeten kwam. (ec)he also on the cinders in all respects out-of the feet came'Charlie Greene showed the more than three thousand spectators that he couldin every respect put his best foot forward on cinders, too,'

(34) Juist omdat hij oneerlijk was, macht ik hem niet latenprecisely because he dishonest was, ought I him not letmerken dat ik hem oneerlijk vond. (ec)notice that I him dishonest found'Precisely because he was dishonest, I should not let him notice that I consid-ered him dishonest.'

In (33), the audience has to do no more than keep their eyes open, in order for themto perceive what Greene does.13 And (34) suggests that were the causee to notice thatthe Speaker found him dishonest, this would affect the causee in an undesirable way,and furthermore than the causee is definitely not actively looking for the causer'sopinion.

Let us now turn to the marking of causees with door, äs in (8) and (35):

(8) Zij wilden Woody daarna door een echtpaar laten adopteren.<ec)

they wanted Woody thereafter through a married-couple let adoptThey wanted to have Woody adopted afterwards by a married couple.'

(35) Want tegenwoordig laten goudsmeden zieh weer graag inspirerenfor presently let goldsmiths themselves again gladly inspiredoor deze klare, strakke stijl. (ec)through this clear, austere styleTor nowadays goldsmiths are happily letting themselves be inspired again bythis clear, austere style.'

As we said above, in the Joor-marked cases causees are categorized äs Instruments.They are not an object of the causer's activity. The purpose of the causer in (8) is notto have some couple adopt a child, it is to get the child a home, and the couple is theInstrument to that end; the only object of the causer's wish is Woody. Accordingly,we typically find indefinite noun phrases äs <ioor-marked causees. In the set of 118sentences in our corpus in which prepositional marking is possible in principle

13 Note that the Initiator here is not a purely physical entity causing the perception, but a volitional one,which motivates the categorization of indirecl causation.

A Veihagen, S Kemmet l Journal qf Pragmatik 27 (1997) 61-32 81

(Table 3), there are 30 indefinite causees; 23 of them (77%) have door; on the otherband, of the remaining 88 definite causees, only 32 have door (36%); this skewingis highly significant.14

Example (35) is a clear case of an instrumental causee which can hardly beexpressed in another way: with zero marking the 'style' would somehow becomepersonified, with the goldsmiths allowing or making it to do something to them,rather than the goldsmiths allowing or making themselves entertain certain ideas (forwhich the style mentioned is an Instrument). Accordingly, quite a number of door-marked causees are also inanimate.

All of these distinctions are in fact parallels to those related to the use of aan, doorand zero in simple clauses, äs is predicted by the analysis proposed in Kemmer andVerhagen (1994). The claim is that causative constructions are modelled on simpleclauses, and from that it follows that aan in causative constructions has the sameeffects äs in simple clauses. Following the analyses of the difference between aanand zero in simple clauses in, among others, Janssen (1976) and Kirsner (1988), it isthen predicted that an αα/ί-marked causee is interpreted äs less affected and moreautonomous than a zero-marked one; äs we have seen, this is indeed the case. Simi-larly, following the analysis of door in Cornelis (1994) and Cornelis and Cuyckens(1995), the effects illustrated in (8) and (35) are precisely those that are to beexpected: door marks minimally affected causal intermediaries with little autonomy.All in all, this in turn further corroborates the general claim from Kemmer and Ver-hagen (1994), that causative constructions are indeed modelled on simple clauseswith only one verb.1"5

6. Conclusion

In this paper we have shown that the two types of analytic causative constructionsin Dutch - those marked with doen and those marked with laten, respectively - canbe best characterized in terms of direct vs. indirect causation. We showed that thisdistinction is closely related to other uses of the verbs, in particular in simple non-causative clauses. This analysis not only accounts for the distribution of the verbs indifferent contexts (especially with respect to animacy of the participants in the causalevent), but also provides insight into the Interpretation of a rieh array of specificcases, in particular when we combine Talmy's theory of force dynamic causationmodels with D'Andrade's theory of the folk model of the mind. Finally, we showed

14 Of the seven aa/i-marked causees, five are definite, and two are indefinite; these numbers are toosmall to show a statistical difierence with the two other sets (dooi- äs well äs zero-markmg). The dif-ference between the latter two itself is statistically very significant. In fact, the latter pattern parallels thegeneral distribution of (m)defmiteness over prepositional phrases suggested by (among others) Kirsnei(1988).'"" Furiher evidence, specific to Dutch, can be provided from word order phenomena: causative con-structions exhibit certain patterns that are restncted to mono-clausal structures. However, we will notelaborate this pomt höre.

82 A Vuhagen <> Kemmei l lownal of Ptagmatic\ 27 (1997) 61-82

that effects of and the restnctions on the occurrence of different case markings foundon causees m Dutch can also be explamed in terms of the analysis of the distinctionbetween doen and laten proposed here, and the general framework for analyzmgcausative constructions put forward m Kemmer and Verhagen (1994).

References

Comne, Bcrnard, 1976 The synlax of causative tonstructions Cioss-language similanties and divcr-genties In M Shibatam, ed , 261-312

Cornehs Louise, 1994 Passing through passives In R Boogaarl and J Noordcgraaf, cds , Nauwebetrekkingen Voor Theo Janssen bij /ijn vijftigste verjaardag, 23-30 Amsterdam/Munsler StichtmgNeerlandistiek VU/Nodus Publikationen

Cornelis, Louise and Hubert Cuyckens, 1995 Gettmg through to Dutch dooi Paper prcsented at the 4th1CLA Conference, Albuquerquc, NM

Croft, William A , 1991 Syntactic categoncs and grammatical relations, The cognitivc oigani/ation ofInformation Chicago and London Universily of Chicago Press

D'Andrade, Roy G , 1987 A folk model of the mind In Dorothy Holland and Naomi Quinn, eds, Cul-tural modcls m language and thought, 112-148 Cambridge Cambridge Umversity Press

Dik, Simon C , 1980 The Dulch causative conslruction In Simon C Dik, ed , Studios m funclionalgrammar, 53-89 London and New York Acadcmic Press

Janssen, Th A J M , 1976 Hebben-konstrukties en indirekt objckt-konslrukties Unpubhshed disserta-tion, Umversity of Nijmegen

Kemmer, Susanne, 1994 Middlc voice, Iransitivity, and the elaboration of events In Barbai a Fox andPaul J Hopper, eds , Voice Form and function, 179-230 Amsterdam and Philadelphia Benjamins

Kemmer, Su/anne and Arie Verhagen, 1994 The grammar öl causativcs and the conceptual structure ofevents Cognitive Linguistics 5 115—156

Kirsner R S , 1988 Prepositional versus 'bare' mdircct objects m the wntten Dutch of novcls and ncws-papers In T Broos, ed Papers from the Third Intcrdisciphnary Conference on Netheilandic Studies,279-296 Lanham, MD Umversity Press of America

Lakoff, George, 1987 Women, fire, and dangerous thmgs Chicago and London Umversity of ChicagoPress

Langacker, Ronald W , 1991 Concept, image, and symbol The cogmlivc basis of grammar Berlin andNew York Mouton de Gruytcr

Renkema, Jan, 1981 De taal van 'Den Haag' Een kwantitatief-stilistisch ondcrzoek naar aanlcidmg vanoordelen over taalgebruik 's-Gravenhage Staatsuitgeverij

Shibatam, Masayoshi, ed , 1976 Syntax and semantics, Vol 6 The grammar of causative constructionsNew York Academic Press

Talmy, Leonard, 1976 Semantic causative types In M Shibatam, ed , 43-116Talmy, Leonard, 1988 Force dynamics in language and cognition Cognitive Science 12 49-100Turner, Mark, 1987 Death is the mother of beauty Chicago and London Umversity of Chicago PressUit den Boogaart, P C , 1975 Woordfrequenties in geschrevcn en gesproken Ncderlands Utrecht Oost

hoek, Scheiternd & Holkema

INSTRUCTIONS TO AUTHORS

The louinal publishe1, ihe following lypes of conlnbutions ( l ) füll lenqth aiticles on all topics of hnguistic pragmatic inteiest,(2) short, free-form sc/i/iM mtended to clanfy theorelical standpomls, iry oul new ideas, report research fmdings or simply toentertam or slimulate discussion among colleagues, (Ί) ι ephei and > ejomdeis to previously published articles, (4) ie\ie» s of booksnot more lhan fwe years old, (5) ie\ie\\ aitnle·; on particularly noteworthy or ihoughl-provoking new books, and (6) brief booknotne\ Three types of issues are published Regulai ;vjnfs containmg articles on differenl aspects öl pragmatic theory andreseaich Spetial Itsuei on lopics of partitular relevance lo currenl discussions in hnguistic pragmatics, and Fötus on Issuev onbroad subject areas of mleresl to different groups of readers Manuscripts accepted for pubhcalion are aulomatically consideredfor mclusion in each type of issue

ManuMiipt·, Manuscripts should be senl lo the Chief Editor, Jacob Mey, Odense Umversity, Campusvej 55, DK-5230 OdenseM Denmark, and should be prepared in accordance with the guidelmes summanzed below Manuiinpls not lonfoimmq \\ilhtitele quidelim v w / / / be tetuined to ihe authois foi levtvon befote betnt> f i t t t he i pioie\\ed Alw the louinal doe^ not tuppoi tduphiale wbmiviiom only manuviipli offeied e\du-,ivel\ to the louinal LOH be tomideiedfoi publuation The pieferred languageot pubhcation is English (allhough exceplions are occasionally made), and manuscripts should meet the louinal ? Englishlanguage Standards (sce below) Authois are expected to provide/m« copies of their conlnbulion, typed ntple-ipaied, with widcmargins, on plam, Standard A4 or U S letter si/e paper Papers should be lyped on one side of Ihe page only, and Ihe pagesshould be numbered conseculively The following Information should appear on the fnst page the füll tille of the conlnbulion,thc author's name and füll addrcss, and an abbrcvialed lille not exceeding 40 characlers includmg spaces An abslract nolexceeding 200 words should be provided, äs well äs a short biographical note not exceeding 100 words per authoi Diskelles arewelcomc only whcn accompamcd by haid copy manuscupts

Citationf QuotatOns should be enclosed m double quotalion marks Smgle ('scare') quoles may be used lo diaw atlention to aparticular Hern in the text Words Irom othei languages, and words mtended lo be especially cmphasized, should be underlmed(itahci/ed) References in ihe texl lo pubhcations should include the author's name, the year of pubhcation, and, if necessary,page numbers, äs m thc following example

Allhough ihis type of convcrsatioiial coordmation has been mvcstigaled from differenl pomls of vicw over the years (Sacks cld l , 1974, Van Dijk and Kmtsch, 1983, Schiftnn, I987a, b) a general theory of language use accounlmg foi it, is still, m Kasher's( I 9 9 l d 129)woids, " "

Bibliotfiapliual Re/eiunc, A relerence hst containmg all works rcfeircd lo in the text and lootnolcs, alphabelically arranged andtyped niple -,paied, should be meluded at the end of the manuscript Authors füll lirst names should be used unless ihe aulhorsthemselves customanly use only miliais References to two or more works by the same author m a smgle year should beaccompamed by a lower-case a b, etc aitei the year of publication, both in the referencc hst and m citalions in the textRelerences to books should include Ihe place of pubhcation and the pubhsher's name and icferences lo articles in Journalsshould include volume and page numbers, äs in the following cxamples

Kasher, Asa, 1991 a Pragmalics and Chomsky's research program In A Kasher, ed , The Chomskyan turn, 122-149 OxfordBlackwell

Kasher, Asa, 1991b On Ihe pragmatic modules A lectuie Journal of Piagmafics 16 381-397Sacks, Harvey, Emanuel A Scheglolf and Gail Jefterson, 1974 A simples! systematics for the organizalion öl lurn-taking tor

convcisaliori Language 50 696-735Schiffrin, Deboiah, 1987a Discouise markers Cambridge Cambridge Umversily PressSchilfrm, Deborah, 1987b Towaid an cmpincal base in pragmatics Language in Society 16 381-396Van Di|k, Teun Λ and Waller Kmtsch, 1983 Slralegies of discourse comprehension New York Acadcmic Press

lOotnote', Footnoles should be lyped tuple tpatcd on a separate page and numbered consecutively They should be äs few andäs short äs possible and should nol conlam bibhographical refeiences except insofar äs Ihese are relevant to the maltersdiscusscd m Ihe loolnoles ihemselves All othcr refercnccs should be given at the appropnate places m the mam text

Γΐίζκιΐ", All tables, hsls figures, i l lus lrat ions and olhei graphics musl be repioducible Originals, and should be prcsenled wi lhappropnaie caplions on sepaiaie sheels, with mslructions mdicating approxnnalely where they are to appear in the text Phrasestiucture irecs, argumemation Schemata, networks, flowcharts, and diagrams should be kepl lo a mimmuin Figures should befreely movable in Ihe lext, and should bc rcierrcd to by nurnber (e g , "Fig 2"), and not by expressions such äs "the figurebelow ', or "ihe ligure above' Illusiialions should be drawn in India ink m large, rcproducible sizc and should be carcfullylelleicd Authol <, M / / / bi chui a,ed if then illuiliatiom ha\e to be le diavai

^vmbol*, Current lype-faccs should be used, and spccial symbols should be avoided äs much äs possible Most Standard logicalsymbols aic prmtablc äs well äs ihe Greek, Cyrilhc, Gothic, and phonelic alphabcts, Chinese characteis, and Japanese lan]ilkana All other symbols from scupls olhei than those above, musl be drawn, in oversize formal, in India ink on separate shccts,lollowing the piocedure outhned for illustralions In läse-, of doubl·, 01 qneition\, authoi s aie adined to ιοηιιιΐι nith the Ednoi\befou <,ubmiltint> iheu mumi\iiipt\

Sfy/t' The Editors reseivc the i ighl lo correcl, or to have corrcctcd, non-native use of English (01 olher languages) The Jownalhas adopted U S English usage äs ils norm (this norm does not apply lo olhcr nalive useis öl English) Also, the louinal isactively commilled lo avoidmg scxisl language äs far äs possible, and to affnmmg and promotmg Ihe use of nonsexislexpressions 11 m doubl authois should consult the Guideltne\ foi Non\e\nt V\e ofLani>uai>e of the American PsychologicalAssociation, published in the Amintan Pivcholoi>itt 30 (1975) 682-684, 01 the Guideliwi öl the Linguislic Society of Amencam the ΖΛ/1 Bulletin (# 135, March 1992)

Authors' bcncfits. Upon pubhcation in the Jouinal of Pia^matu \, authois receive (1) 50 icprints of thcir connibulion free ofCharge, and (2) a 30% discount on all Elscvicr Science B V books

US mailing noticc - Journal of Pragmatits (0378 2166) is published monthly by Elsevicr Science (Molenwcrf l , Postbus 2l l1000 AE Amsterdam) Annual subscnplion pnce in ihe USA US $ 568 00 (US $ pncc vahd in North Ceniral and Soulh Amencaonly), includmg air speed dchvery Sccond class postagc paid at Jamaica, NY 1143 USA POSTMASTERS Send addresschanges to Journal of Pragmalics Publication Expeditmg, Ine, 200 Meacham Avenue, Elmont, NY 11003 Air f re ighl andmailing in thc USA by Pubhcation Lxpcdiling

JOURNAL OF PRAGMATICS• EDITORSChief Editor: J.L. Mey, Odense Umversity,Campusvej 55, DK-5230 Odense U, DenmarkExecutive Editon R.W. Janney, Universitätzu Köln, GermanyEditore: C. Catfl, Universitä diGenova, Italy,T. Frethelm, University öl Trondheim, Norwayand H. Haberland, Umversity ölRoskilde,DenmarkReview Edltor: M. Dolitsky, InstitutUniversilaire de Formation des Maitre(IUFM), FranceSqufe Editor B. Fräser, Boston University, USAAssociate Editors: L. Coleman, Universityöl Maryland, USA, f. Kiefer, HungarianAcaderny of Sciences, Hungary,J.-O, Östman, University öl Helsinki, Finlandand J. Versteueren, University olAntwerp,Belgium

• AUDIENCELinguists, Sociologists, Psychologists,Anthropologists, those involved in ArtificialIntelligence, Philosophers interested inpragmatics, semantics, languageacquisition, universals of language,computational linguistics, sociolinguistics,speech act theories, discourse analysis etc.

• ABSTRACTED/INDEXEDINArts and Humanities Citation Index,Cambridge Scientific Abstracts, CurrentContents/Arts and Humanities, Gurren!Contents/Social and Behavioral Sciences,INSPEC Information Services, LanguageTeaching, Linguistics Abstracts, Linguisticsand Language ßehavior Abstracts (LLBA),Social Sciences Citation Index, SocialPlanning/ Policy and DevelopmentAbstracts (SOPODA), SociologicalAbstracts, The Philosopher's Index.

• ©1994 SUBSCRIPTIONS

Volumes2l-22(in 12issues)ISSN 0378-2166Please contact the publisher for theSubscription Price

An InterdisciplinaryMonthly of Language

StudiesCommenced publlcation 1977

ELSEVIERSCIENCESIMPRINT: NORTH-HOLLAND

• AIMSANDSCOPELinguistic pragmatics has baen able toformulate a number of queslions (hat areessontiai to our understanding of language äshumans' main Instrument of "nafi/ra/"and*soc/e/a/"interaction. ßy providing possibletheoretical foundations for the study oflinguistic practice, Imguistic pragmalics hashelped to increase our Knowledge of humaninteraction. The Journal öl Pragmaticsidentifies with the above general scope andaims of pragmatics. In addition, the Journalwill attempt to narrow the distance betweensuch fields äs sociolinguistics,psycholinguistics, computational linguistics(including the Information sciences, artificialintelligence research and, in general, thetheory and practice of machine-humaninteraction), applied linguistics (including thetheory and practice of language teaching andlearning), psychiatry (induding the study ofSchizophrenie speech). The Journal will alsoendeavour to cover the "s/iady" areasbetween these fields and between linguisticsand the neighboring disciplines. The Journalof Pragmatics welcomes both thosecontributions originating in linguistics properand those taking neighboring, related fieldsäs their point of departure.

Recently the Journal oi Pragmatics alsohas begun to publish focus-on-issues onbroad subject areas of general interest todifferent groups of readers. The first issue ofthis new type was "Pragmatic Paradigms" in1992. Further issues are on "Genderand"Conversational Strategies" (1993)."Special Issues" devoted to a single topic areregularly published: Australian Contributionsto Discourse Theory in 1989, Politeness andSelected Papers Irom the InternationalPragmatics Conference, Antwerp 1987(1990), Syntax and Pragmatics (1991),Cognitive Perspectives öl Language Usern1991, Merjeclions (1992) and Pragmatics inAustralia and New Zealand (1993),

Send this form to your regulär supplier or toELSEVIER SCIENCE B.V.Attn. Marlene HamersP.O. Box 1031000 AC AmsterdamThe NetherlandsFax: (+31-20) 5862 616

Customers in the USA/Canada, send toELSEVIER SCIENCE INC.Attn.JudyWeislogel,P.O. Box 945Madison Square Station, New York,NY 10160-0757

Q Please send me a FREE inspection copy ofJOURNAL OF PRAGMATICS

Q Please enter my 1994 Subscription toJOURNAL OF PRAGMATICS

O Yes, l would like to be on your Mailing List:(mailing code 08736)

ORDER FORMQ Payment Enclosed: (Bank Draft / Eurocheque / International Money Order /

Postal Cheque / Oflicial Purchase Order Form)Q Please send me a pro forma invoiceQ Charge my credit card:

Q Access Q Eurocard QVisa Q American Express Q MasterCard

Card No Valid Until :....Name Initials

Signatare Date..Vat Registration No

The Dulch Guilder (D/1.) price Is delimtive. Cuslomers in Ihe European Community should add theappropriate VA T rate available in their counlry to the above prices. In New York State, p!ease addapplicable sales lax, Prices are subject to change wühoul prior notice.